Thoma & Vista
Simple question here: I'm curious about what I feel like is a logical flaw in the Thoma / Vista investing strategy. This applies to all of software PE, but Thoma and Vista are the biggest so I think it's applicable especially to them.
Let's say you're buying out a software company. You think it's going to grow tremendously, a lot of room for operational improvement, whatever. And you value it at 10x Revenue -- not uncommon in the industry as far as I can tell. This might give you a ridiculously high implied EBITDA multiple, but that doesn't matter -- you're entering on revenue, and you're confident about the business. Maybe you'll exit on EBITDA, maybe not, whatever.
The issue is that banks, when they decide how much financing to give you, don't care how you valued the company. They look (as far as I know) at EBITDA-based leverage multiples. They don't care if you valued the company at 10x Revenue because of x,y,z reason. They'll only give you a specific amount based on EBITDA. As far as I know, for a bank to decide to lend a PE firm more than 15x EBITDA is already pretty rare. And this is amplified by the fact that the cost of debt is so high right now that most sponsors aren't going to want that much debt, these companies are probably riskier (if they're not being valued on EBITDA, they're probably not that profitable / FCF-generating), and so many other things.
So, how does the ultimate financing package look? If the implied EBITDA multiple is something crazy, like 200x, and Goldman Sachs or BofA or whoever is only willing to give you 10x debt financing, does that mean you're effectively putting in 95% equity? Doesn't that kill the entire point of an LBO? Are these companies effectively just cutting pure equity checks for (almost) the whole enterprise?
This obviously doesn't apply to growth investments where you're looking at no leverage. Maybe I have no clue what I'm talking about and banks will buy into your crazy vision and let you have 120x EBITDA worth of debt. But that seems unlikely. I'm probably missing something very simple here. Thanks in advance for the responses.
Yes, you are missing something very simple. It's called ARR lending, which has debt / ARR covenants. For most bank lenders, the covenants convert to debt / EBITDA 18-36 months into the term, but unitranche "private credit" or "direct lenders" can offer more flexibility.
Makes sense. That is pretty simple. So how do lenders account for the additional risk that they’re taking on by writing ARR-based financings? Obviously it’s not revenue that accounts for a company’s ability to service debt, but cash flow. So a company that is generating limited FCFs but with a strong ARR profile is just not going to be able to service debt as well on a $ basis. Do banks charge higher rates for ARR financings? Stricter covenants?
Correct, you do not.
Great, thanks so much for being helpful! Next time I don’t know something, I’ll just close up my laptop and literally just fuck my own face instead of asking people who may be more knowledgeable.
One of the best movies ever and Tom's best role by far
The assumption you are making is that the software businesses a Thoma, Vista, Hg, etc. buy don't generate much (if any) EBITDA / cash flow. Most of these businesses are already generating a ton of EBITDA / cash and can therefore sustain significant amounts of leverage - of course as you say in a high interest rate environment these businesses are a bit more constrained but that is true for all industries. On a relative basis, software is likely still one of the better places to deploy / underwrite for direct lenders / banks and at higher avg. leverage multiples in today's environment... would you rather lever up a software business 6x today (high gross retention, pricing power, high and reliable cash conversion / limited capex) or an industrial company 4x (inflation pressure, commodity exposure, likely procyclical, possibly lumpy capex requirements) where in 24 months you could be 7-8x levered on businesses that may only be worth HSD and may run into liquidity issues due to fixed cost deleveraging?
I think your theory applies least (not most) to the Thoma's, Vista's, Hg's of the world and perhaps more as you move down market and especially into VC-backed businesses. The idea that all tech / software companies are growth at all cost with no ability to generate cash probably stems from the fact that a lot of high-profile VC-backed (and now public) companies have gone down that path because it's been rewarded over the last 10+ years by VC / GE firms and the public markets but those businesses are typically not the key targets of software- / tech-focused buyout firms. And when they do buy businesses that look more a bit more like these VC-backed companies they will as you say probably overequitize at the outset and may add leverage over time as the business matures and can sustain leverage.
I interned at Vista, and the above poster actually isn't accurate. Many of the companies Vista buys don't product cash flow, this is how they add value and juice returns. Robert Smith has talked about this extensively on podcasts and has essentially created the to ability finance software companies. Robert Smith was the head the Software IB at GS. He realized that ARR is a intangible asset and is actually more attractive to bank lenders than CF/EBITDA, since these are 3-5 year guaranteed contracts already agreed upon by companies. When you combine that with the fact that Software businesses have 80-90% gross margins and capex ~1% of revenue it makes lending to these software businesses actually a no brainer. This makes bank lenders comfortable because the downside case is baked in if the business does 0 growth because they have guaranteed revenue for the next few years, and is all upside if the business continues to add new seats. Vista will enter an investment using an ARR multiple as the basis for debt financing, but exit the investment on a EBITDA multiple basis. An example of this investment strategy is their investment Marketo which was bought by Adobe. Marketo was not profitable and Vista turned it profitable and sold the business for like 3x MOIC in like 2 years
Another thing is that Vista uses less leverage than the normal PE firm. It is very common to see Vista putting in 50-70% equity into a business rather than 50-70% debt. If you look at Vista's first fund, all their deals was 100% equity.
Vista and Thoma also have different strategies. Vista is more operationally focused, whereas Thoma's strategy is closer to Audax's buy and build strategy.
I respectfully disagree that my post is inaccurate. I will concede that Vista is probably the worst example out of the names mentioned to illustrate my point because of the focus on their "value creation playbook" which is all about increasing profitability in target companies... this clearly allows them to look at companies with different profiles, including businesses with limited profitability on day 1 where they may resort to ARR financing and over time move towards more "traditional" financing methods as they help drive additional profitability. That said, Vista does buy businesses that generate plenty of EBITDA out of the gate as well (Citrix take-private / combo with TIBCO in 2022 is an example of a highly profitable business, EngageSmart, etc.) where they can leverage "traditional" lending sources like direct lenders / syndicated debt markets to raise debt on day 1. A more accurate statement would have probably been "many (not most) of the business are already generating a ton of EBITDA / cash". The fact that Vista's first fund was all equity, if anything, just proves my point that OP's thesis of software companies can't raise debt applies a lot more to the middle market (Vista Fund I was <$1bn in commitments) where you run into software businesses that are less mature and have not achieved the scale to achieve significant profitability.
The other names mentioned (Thoma, Hg, but also other MFs like Bain Capital, H&F, etc. or Francisco Partners, Marlin that play / dabble in software) buy plenty of businesses that are already profitable and are both valued and levered based on "traditional" EBITDA metrics. That doesn't mean every single one of their software deals fits that profile, which is why I mentioned they will typically overequitize initially in situations where businesses are not yet / not very profitable.
The point is more that OP was asking about whether there is any debt to be raised with software companies and the answer is yes, either via ARR related financings OR because lots of the software businesses targeted by PE firms actually produce enough cash to sustain meaningful amounts of leverage. This is also borne out by data from the direct lending group at my firm, which did lots of deals with software-focused funds that were always on the "limit" of what you would see in terms of total debt quantum (6-8x leverage based on PF Adj. EBITDA, plus HoldCo debt on top in many cases).
Tl;dr ARR-based lending is a thing. In the 2020-2021 heyday market was 2.0-2.5x ARR for reasonable terms; in current market it's closer to 1.5x ARR. In '20-'21 it was also pretty common to see these facilities PIK for the first couple of years so that the sponsor could have time to drive profitability; in current market that is much less common, and you're seeing sponsors pre-fund a year of interest coverage.
A logical follow-up here is that if a company’s being valued at 10x revenue, and ARR lending sets a limit at 1.5-2x, you’re still looking at about an 80% equity based deal, which is pretty high for an LBO (usually looking at around 40%). Does this mean that for the most part, debt as a % of total purchase price with these ARR-based valuations / lending packages are much lower, possibly 1/2 of what you’d see if you’re doing a traditional LBO of a company in some other industry? And at that point, why even use debt at all?
Obviously there are a lot of variables here — apples to oranges doesn’t do us any good, so let’s take 2 situations, both which pose us with a different issue:
Situation 1 (Hypothetical Software company with a healthy margin profile, just as healthy as a hypothetical industrials company):
Software company with 100 Revenue and 20% EBITDA margin valued at 10x Revenue and therefore implied 50x EBITDA. Industrials company with the same financial profile except valued at 10x EBITDA. (Pretty common valuation profiles to see). Issue: In the case of the software company where you use ARR lending at a max of 2x, you have to cut a 80% equity check. In the case of the industrials company where you use EBITDA lending at a (arbitrary but somewhat standard) max of 6x, you’re looking at a very standard 40% equity check. 80% equity for the SW buyout — at that point, why do an LBO at all?
Situation 2 (Software company with a weaker margin profile than the industrials company):
Software company with 100 Revenue and 5% EBITDA margin valued at 10x Revenue and therefore implied 200x EBITDA. Industrials company with the same profile as in Situation 1, with 100 Revenue and 20% EBITDA margin valued at 10x EBITDA. (Again, somewhat common profiles). In the case of the software company where you use ARR lending at a max of 2x, you hardly have enough cash flow generation to make interest payments, so your cash generation will go down an undesirably large amount (possibly negative), and returns will be mediocre here too— again, why use debt at all if it’s just hurting you? Sure, the argument is that you’re predicting the company will grow astronomically — maybe you’re PIKing the debt so you don’t have to worry about interest — but again, why go through all this just for a very marginal benefit that you get with a 80/20 equity/debt split rather than 100/0?
To summarize the separate issues:
Situation 1. If the software company is indeed generating healthy FCFs, doesn’t the fact that it’s valued so much higher mean that you can’t really raise that much debt for the LBO for it to make any more than a marginal returns improvement compared to a pure equity buyout?
Situation 2. And if the software company has low FCF generation already, what’s the point of putting debt on it and bringing that FCF generation even lower with interest payments (as well as a whole slew of new issues like meeting covenants, etc.), just for the sake of this aforementioned marginal returns improvement?
Running LBOs has taught me that determining the amount of leverage to finance a transaction is sort of like the elementary school problem of finding the maximum area you can create with a rectangle given a certain amount of length for the rectangle’s sides. Let’s say you’re given a maximum length of 16. You can try 2x6, that gives you 12. You can try 5x3, that gives you 15. Ultimately you realize that 4x4 (the square) is always going to give you the largest possible area. In an LBO, too little debt brings you returns that are too low because you’re not maximizing the potential of the FCFs available for paydown. Too much debt hurts the FCFs too much and reduces the paydown potential itself. There’s a perfect balance somewhere there which comes with trial and error.
For a software business valued at a 10-20x revenue multiple, it seems like this “sweet spot” will just never get you high enough returns, at least in theory — because you just can’t raise enough debt to cover 60-70% of the purchase price, which is usually where that sweet spot is. If you do model out a 20x revenue company with 60% debt, you’re either taking on way more debt than is realistic (nobody’s going to give you 12x ARR), or (more frequently), you taking on so much debt hurts your cash flows too much to make a positive impact on returns.
Nesciunt dolorem sit possimus itaque est labore dolores et. Quo nostrum earum molestiae ipsa assumenda velit consectetur.
See All Comments - 100% Free
WSO depends on everyone being able to pitch in when they know something. Unlock with your email and get bonus: 6 financial modeling lessons free ($199 value)
or Unlock with your social account...